Neurox Media Network: idolforums.com | aikenforums.com | IDF Webmail  
Neurox Media's American Idol 10, 9 & 8 / So You Think You Can Dance 8, 7 & 6 / Glee / X Factor Forums & Message Boards
Neurox Advertising   The banner image below is an ad.

Welcome Guest ( Log In | Register - It's quick & free! | Resend Validation )    

4 Pages V  1 2 3 > »    ( Goto last unread Go to first unread post )
( Star this topic Star this topic | View starred topics )    Reply to this topicStart new topic
> How Would Siobhan Fared with Jennifer Lopez & Steven Tyler?
Rohm
post Sep 30th 2010, 1:20 AM
Post #1



**********

Group: Members

Posts: 4,120
Gender: Male
From: Long Island, New York



If Jennifer Lopez & Steven Tyler were both AI9 judges:

How Would Siobhan Fared with Jennifer Lopez & Steven Tyler?

Your thoughts, please.

Thanks.

This post has been edited by Rohm: Sep 30th 2010, 1:22 AM


--------------------
My AI8 fav: Adam Lambert
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
NX AdBot™
post Sep 30th 2010, 1:20 AM


NX Advertisement


Group: Marketing

Posts: I've been known to spam...
Gender: Not telling :P


Subscribe to NX to remove this advertisement post. Help us pay for our servers! Click for subscriber benefits.

 
 
Go to the top of the page
Tichwow
post Sep 30th 2010, 2:31 AM
Post #2


 
******

Group: Members

Posts: 832
Gender: Female
From: Hollywood
Fav. AI5: Elliott Yamin



I think this thread has already been done.


--------------------
IPB Image
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
blacklightposter
post Sep 30th 2010, 9:46 AM
Post #3


 
******

Group: Members

Posts: 543
Gender: Male



QUOTE(Tichwow @ Sep 30th 2010, 3:31 AM) *

I think this thread has already been done.
I think it was done re: Steven Tyler.

And a JLo thread was created and going, but a mod closed it as not Sio-related. (Coulda easily been made Sio-relevant just by a small touch-up of the title, but oh well.)


What Rohm is doing here is a -- how would Siobhan have fared under the whole new judging line-up(?) thread -- which is a little different.

How would a Randy-JLo-Steven Tyler judging panel have treated her?


I'm assuming that all of those judges would've been honest and not had some agenda they were trying to execute. And if that's true, then I think they'd have praised Siobhan and promoted her as clearly the best of the season.

But that of course isn't the same thing as saying she'd have won. The judges have a significant impact, really only on the sheep-like viewers, but there are many of those. They obviously do not control voting, though, and even supported across the board by a fair judges panel, Siobhan would still have faced AI's f--ked up voting system.

She'd still have been eventually crushed by the mountain of indifference from Boy-Crazy Powertexting Tween Girl Nation, which completed their coup of AI a few years ago and now completely decide the final outcome of it year in and year out.

With fair judging, judges would have promoted her to the viewers as the best of the season, and she'd have made it, maybe, to, say, 3rd place. Or who knows, maybe she could've even peeled off more of Crystal's older-female voter base and even gotten 2nd place... but I think 3rd is more likely.

This post has been edited by blacklightposter: Sep 30th 2010, 9:59 AM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Sep 30th 2010, 10:37 AM
Post #4


 
****

Group: Members

Posts: 146
Gender: Male



It's plausible that AI & AT&T are in the middle of a multi-year contract that prevents AI from making any changes to the voting system.

However... I have to believe AI is aware of the fact that the audience is becoming sensitive to the appearance of a math problem when looking at the Nielsen Ratings and the vote totals Ryan reports.

Twenty million viewers voting thirty six million times over 120 minutes (making 8 second calls) would never cause the phone system in this country to go into terminal busy signal mode if tens of millions of votes were not being tossed away in one way or another.

______

I forgot to mention text messages. Is AT&T relaying accurate numbers to AI's vote management company as AT&T would be getting the hard data first.

All those text messages, maybe 40% to 60% of the vote totals, were cast by people that never called a phone number.

______

So no... unless the voters wise up and realize that "x" number votes are counted as good votes and "x+1" is not counted at all because its considered a power vote (using technology to vote more than the "average voter") nothing will change, or would have changed.



--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Rohm
post Sep 30th 2010, 12:05 PM
Post #5



**********

Group: Members

Posts: 4,120
Gender: Male
From: Long Island, New York



Problem is: AT & T is a major sponsor of AI and may have an existing long-term contract.

Thus, AI & AT&T are linked in a major way for quite some time..

Doubt that AI will be willing to rock the boat while AT& T is a major sponsor.

AT &T has no incentive to leave AI due to the big phone revenues from texting fees.

Bold (utopian) alternative for AI:

1. drop AT& T as major sponsor upon contract extension, then
2. revamp/limit the voting system

This post has been edited by Rohm: Sep 30th 2010, 12:06 PM


--------------------
My AI8 fav: Adam Lambert
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
blacklightposter
post Sep 30th 2010, 4:20 PM
Post #6


 
******

Group: Members

Posts: 543
Gender: Male



QUOTE(Rohm @ Sep 30th 2010, 1:05 PM) *

Problem is: AT & T is a major sponsor of AI

and may have an existing long-term contract.


Agree with the first half of that.

But as for the 2nd half, there are all sorts of conditional and contingent provisions in business contracts, put there to protect the company. And I can just about guarantee that if AI concludes that their arrangement with AT&T is hurting their show and therefore their own business, that they have the means to get out of either the entire arrangement or at least alter the damaging parts of the contract.

AI surely understands that this voting problem is a very large contributing factor to the general public's declining interest in the show... and therefore their declining ratings, the show's revenues, and even its long-term prospects for survival. Their very well-paid lawyers didn't let them sign anything with AT&T that would require one going venture (AI) to destroy itself so that one stream of income to another going venture (AT&T) stays undiminished (and only for awhile at that). Nor would any court force them to let their ratings self-destruct for such a reason.

QUOTE(Rohm @ Sep 30th 2010, 1:05 PM) *

Doubt that AI will be willing to rock the boat while AT& T is a major sponsor.

AT &T has no incentive to leave AI due to the big phone revenues from texting fees.


Agree with the 2nd part of that, but not the first.

What AT&T is gaining through the AI tween viewer powertexting phenomenon is customers, and especially sales of its $20/month unlimited texting options. It has sold lots of those, believe it or not, specifically because a small army of tweens passionately want to be able to personally cast 1000 or more votes in two hours for their fave cute guy singer. Sale of that option adds up to good money. And yes AI does get lucrative sponsorship in return.

But what AI has to decide, and probably now, is if this large of a flow of AT&T money this year really is worth allowing viewership of the show to continue to decline, and the show to possibly last only a few more years? Is that the kind of business decision they should be making? Because, while fixing the voting system problem alone will not save the show, it still must be done if the show is to be saved. It is a "necessary but not sufficient [by itself]" step to slow or stop AI's decline.

Or is it smarter to let go of some of that AT&T money today (or potentially even all of it, if AT&T is offended enough.. but replace them with other sponsors), and limit votes cast per phone in order to fix a very serious problem hurting viewer interest in their show?

My decision right now, were I AI management, would be the second of those choices.

QUOTE(Rohm @ Sep 30th 2010, 1:05 PM) *

Bold (utopian) alternative for AI:

1. drop AT& T as major sponsor upon contract extension, then
2. revamp/limit the voting system


A little bold, maybe, but not so utopian. I'm positive a vote-per-phone limit could be accomplished without a lot of difficulty, and it might not even mean losing AT&T entirely as a sponsor. Millions of people still need to phone in somehow. It might just mean a little less enthusiastic and generous a corporate sponsor.



--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
A Song To Sing
post Sep 30th 2010, 5:05 PM
Post #7



*********

Group: Members

Posts: 2,284
Gender: Male



As I've outlined before, I think the judges' decision to use the save on Mike was what ruined Siobhan. Had they not saved him, either Katie or Tim would have made it to the Top 6, and I can't imagine either one getting more votes than Siobhan. I also think Siobhan would have gained a lot of momentum the following week (Sinatra week) since there were so many weak performances. I give her a 50% chance of winning it all if the save is never used.

This post has been edited by A Song To Sing: Oct 8th 2010, 11:21 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Sep 30th 2010, 5:27 PM
Post #8


 
****

Group: Members

Posts: 146
Gender: Male



QUOTE(A Song To Sing @ Sep 30th 2010, 6:05 PM) *

As I've outlined before, I think the judges' decision to use the save on Mike was what ruined Siobhan. Had they not saved him, either Katie or Tim would have made it to the Top 6, and I can't imagine either one getting more votes than Siobhan. In addition, I think Siobhan would have gained a ton of momentum the following week (Sinatra week) since there were so many trainwreck performances. I give her a 50% chance of winning it all if the save is never used.


Brian your post implies that Sio didn't have enough votes to stave off elimination. I contend she had TOO MANY VOTES (that were disqualified).

Just take out a calculator and see how 20M viewers fits into 36M votes. Even if 5% of the viewers vote (1M voters) did they stop voting at 36 times via phone and text over 2 hours?

OK so you're answer is yes... 1M out of 20M actually vote and they stopped on average at 36 votes in 2 hours.

What does that say about the people that voted 400/500/800/1200/2000/3000/4000 times?

It says their votes went POOF!

Vote for the worst with a base of 200K members could have cast 20M votes themselves by voting 100 times.

Siobhan is/was a victim of the Vote Management Company and their directive to keep "power voters" from unfairly influencing the outcome. And I'm OK with that... just tell us what the magic number is when your vote total becomes "1 vote too many".

The "save" meant nothing and still means nothing.

This post has been edited by jackxyz: Sep 30th 2010, 5:30 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
A Song To Sing
post Oct 1st 2010, 1:11 PM
Post #9



*********

Group: Members

Posts: 2,284
Gender: Male



Jack, I think you are rushing to conclusions without adequate knowledge of the underlying statistical patterns. I don't have any hard evidence, but I suspect that the vast majority of Idol voters never cast more than a few votes each week. Although power voters have a big impact on the results, they are but a small minority of the voting audience. They don't contribute as much to the average number of votes per voter as you might think.

Let me illustrate my point with a few hypothetical examples:

If 80% of voters cast 10 votes, and 20% cast 500 votes, the average number of votes is 108.

If 90% of voters cast 10 votes, and 10% cast 500 votes, the average number of votes is 59.

If 95% of voters cast 10 votes, and 5% cast 500 votes, the average number of votes is 34.5.

If 98% of voters cast 10 votes, and 2% cast 500 votes, the average number of votes is 19.8.

The actual frequency distribution is going to be a lot more complicated than this, but my point still stands. The average number of votes per person is going to be well under 100. I suspect that it is in the 10-20 range. You are probably underestimating how many people contributed to that 36 million vote total during Shania Twain week, and how many of those were nothing more than casual voters (particularly country music fans).

I think Siobhan was simply in a lot more danger than any of us realized. Female rockers seem to have an expiration date on this show, Carly Smithson being a perfect example of this. Like Siobhan, she was eliminated in 6th place after giving a stellar performance that was praised by all the judges. She was also never in the Bottom 3 until the night of her elimination. It would appear to me that casual voters always gravitate to the more "mainstream" contestants as the show progresses. That's why someone like Siobhan can appear to be safe for weeks on end, then suddenly find themselves voted off the show.

This post has been edited by A Song To Sing: Oct 1st 2010, 4:16 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Bonbonconvert
post Oct 1st 2010, 4:14 PM
Post #10



*******

Group: Members

Posts: 1,439
Gender: Male
From: Kalispell, Montana
Fav. AI9: Siobhan Magnus
Fav. AI7: David Archuleta
Fav. AIXIII: Jena Irene



QUOTE(A Song To Sing @ Oct 1st 2010, 12:11 PM) *

Jack, I think you are rushing to conclusions without adequate knowledge of the underlying statistical patterns. I don't have any hard evidence, but I suspect that the vast majority of Idol voters never cast more than a few votes each week. Although power voters have a big impact on the results, they are but a small minority of the voting audience. They don't contribute as much to the average number of votes per voter as you might think.

Let me illustrate my point with a few hypothetical examples:

If 80% of voters cast 10 votes, and 20% cast 500 votes, the average number of votes is 108.

If 90% of voters cast 10 votes, and 10% cast 500 votes, the average number of votes is 59.

If 95% of voters cast 10 votes, and 5% cast 500 votes, the average number of votes is 34.5.

If 98% of voters cast 10 votes, and 2% cast 500 votes, the average number of votes is 19.8.

The actual frequency distribution is going to be a lot more complicated than this, but my point still stands. The average number of votes per person is going to be well under 100. I suspect that it is in the 10-20 range. You are probably underestimating how many people contributed to that 36 million vote total during Shania Twain week, and how many of those were nothing more than casual voters (particularly country music fans).

I think Siobhan was simply in a lot more danger than any of us realized. Female rockers seem to have an expiration date on this show. Carly Smithson is a perfect example of this. Like Siobhan, she was eliminated in 6th place after giving a stellar performance that was praised by all the judges. She was also never in the Bottom 3 before she was eliminated. It would appear to me that casual voters always gravitate to the more "mainstream" contestants as the show progresses. That's why someone like Siobhan can appear to be safe for weeks on end, then suddenly find themselves voted off the show.


Hmmmmm... I'll play along

In 2008 the average household size was 2.62, my family used to vote about 5 times a week, so if everybody else was the same there should be ~38,000,000 votes cast (20,000,000/2.62 x 5 = 38,167,938.93). Now this year I really got into voting for a certain someone even if it was 1% of HOUSEHOLDS that voted 100 times that vote total would balloon to ~58 million total votes.

Of course, maybe most of the viewers don't vote at all, if that was the case maybe my numbers don't mean a thing.

I don't think Siobhan has that huge a fanbase, devoted yes, but small compared to someone like Lee. Check out vote fair link, she was at the top for the longest time, but now she's just fourth (still #1 in first choice votes).

This post has been edited by Bonbonconvert: Oct 1st 2010, 4:20 PM


--------------------
AI 6: Melinda, Jordin
AI 7: David A, David C, Syesha
AI 8: Allison, Adam
AI 9: Siobhan, Crystal, Casey, Lilly
AI 10: Casey, Thia, Haley, Pia, Lauren
AI 11: Jessica, Elise, Hollie, Erika, Joshua, Skylar
AI 12: Candice, Amber, Kree, Angie, Devin
AI 13 Jena, Alex, Caleb, Malaya, Jessica
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Oct 1st 2010, 5:24 PM
Post #11


 
****

Group: Members

Posts: 146
Gender: Male



The Vote Management Company responded to the question in a way that implies that power voting can happen with texting 5 votes at a time...

"As you correctly assume, I cannot provide you with the limit of votes before we reserve the right to remove votes. All I am able to state is that if you are using standard means to vote then you should have no issues and all votes will be cast and counted. Should you be using an automated system, you increase your chances of having votes removed. One thing you should also be wary of is sending multiple texts in a single go. Some handsets allow you to put many recipients (in this case it would be the same 4 digit shortcode) in the ‘to’ field and just send one text. While some handsets only allow you to send up to 10 at once, others allow far greater than that. In those instances there may be a chance that they would be considered to be blocks of votes and they may be miscounted.

We encourage people to vote as often as they like using standard means to vote. Anything that abuses or takes advantage of technology to cast large blocks of votes, runs the risk of being disallowed.
---------

Hi again

I’m afraid I cannot give out specifics. Ultimately it comes down to the overall number of votes cast in the 2 hour window. Once someone reaches a certain threshold we then look at the patterns of voting to determine what methods are being used to cast the votes (ie, what blocks). Someone simply sending one message at a time for the whole 2 hours would never get to the threshold, while someone sending multiples (such as 5 at a time) could get to that threshold if they voting constantly for the 2 hours. Sending multiples is taking advantage of the technology and so, while you may never get to the threshold and have any votes discounted, you increase your chances of us removing those votes which exceed the threshold.

I’m sorry I cannot be more specific.

Thanks

Sandy"
[color=#3333FF]

_______

Look... the answer to the question is 36 million votes and 20 million viewers.

So what is the question?

How do 20M viewers cast 36 million votes in 120 minutes?

The answer boils down to certain numbers of potential combinations.

20M ÷ 36M votes = 1.8 votes per viewer

Or

If only 10% of viewers actually vote then 2M ÷ 36M = 18 votes per viewer who took the time to vote.

5% of viewers take the time to vote = 1M ÷ 36M = 36 votes per viewer who took the time to vote.

See a pattern?

2.5% of viewers vote (97.5% don't bother to text or phone in a single vote) =
500,000 voters ÷ 36Million votes = 72 votes per viewer over 120 minutes = 1 vote every 100 seconds.

Even in the worst case scenario where only 2.5% of voters vote over 2 hours that means the average voter stops at 72 votes.

No way is the vote management company acknowledging the votes from people voting many hundreds and sometimes thousands of times.

This post has been edited by jackxyz: Oct 1st 2010, 5:25 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Oct 1st 2010, 5:31 PM
Post #12


 
****

Group: Members

Posts: 146
Gender: Male



* Would the primary sponsor spend multi-millions in advertising if the average text voter stopped after voting twice or twenty times?

* For those who want to believe "most people don't vote", a large segment of the viewing audience is the 14-28 year old market. The market that the primary sponsor targets for unlimited text plans. They don't vote?

* The phone system in the United States would not see a busy signal problem if only 500,000 viewers were making those "long" 10 second phone calls. (or 1 million or 2 million... they are 10 second calls). Nor would AI & their local phone company have an issue with the 866 trunk lines handling those phone calls.

* There is data from the major cellular carriers (besides the primary sponsor) that suggest that they handle an extra 10 - 20 million calls each during AI voting time periods.

* With text voting accounting for 40/60% of the voting traffic, other cellular carriers accounting for another 20-40M calls, plus 866 calls over traditional land lines, its hard not to see a problem with the vote tally as reported the following day.

______

How you choose to explain away what appears to be a problem is your choice. Looking at the numbers and other factors... "power votes" being tossed at least suggests an honorable reason the numbers don't jive.

+++++

Hi again

I’m afraid I cannot give out specifics. Ultimately it comes down to the overall number of votes cast in the 2 hour window. Once someone reaches a certain threshold we then look at the patterns of voting to determine what methods are being used to cast the votes (ie, what blocks). Someone simply sending one message at a time for the whole 2 hours would never get to the threshold, while someone sending multiples (such as 5 at a time) could get to that threshold if they voting constantly for the 2 hours. Sending multiples is taking advantage of the technology and so, while you may never get to the threshold and have any votes discounted, you increase your chances of us removing those votes which exceed the threshold.

I’m sorry I cannot be more specific.

Thanks

Sandy"
[color=#3333FF]


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
blacklightposter
post Oct 1st 2010, 7:01 PM
Post #13


 
******

Group: Members

Posts: 543
Gender: Male



QUOTE(jackxyz @ Oct 1st 2010, 6:31 PM) *

* Would the primary sponsor spend multi-millions in advertising if the average text voter stopped after voting twice or twenty times?

* For those who want to believe "most people don't vote", a large segment of the viewing audience is the 14-28 year old market. The market that the primary sponsor targets for unlimited text plans. They don't vote?

* The phone system in the United States would not see a busy signal problem if only 500,000 viewers were making those "long" 10 second phone calls. (or 1 million or 2 million... they are 10 second calls). Nor would AI & their local phone company have an issue with the 866 trunk lines handling those phone calls.

* There is data from the major cellular carriers (besides the primary sponsor) that suggest that they handle an extra 10 - 20 million calls each during AI voting time periods.

* With text voting accounting for 40/60% of the voting traffic, other cellular carriers accounting for another 20-40M calls, plus 866 calls over traditional land lines, its hard not to see a problem with the vote tally as reported the following day.

______

How you choose to explain away what appears to be a problem is your choice. Looking at the numbers and other factors... "power votes" being tossed at least suggests an honorable reason the numbers don't jive.

+++++

Hi again

I’m afraid I cannot give out specifics. Ultimately it comes down to the overall number of votes cast in the 2 hour window. Once someone reaches a certain threshold we then look at the patterns of voting to determine what methods are being used to cast the votes (ie, what blocks). Someone simply sending one message at a time for the whole 2 hours would never get to the threshold, while someone sending multiples (such as 5 at a time) could get to that threshold if they voting constantly for the 2 hours. Sending multiples is taking advantage of the technology and so, while you may never get to the threshold and have any votes discounted, you increase your chances of us removing those votes which exceed the threshold.

I’m sorry I cannot be more specific.

Thanks

Sandy"
[color=#3333FF]


Jack, a few questions (...then a string of observations):

Aren't those news reports and anecdotal stories from earlier, such as during AI high points? I never followed all of that closely, but didn't votes in a finale or two hit crazy numbers like well over 100 million? I seem to recall something like that during the Adam-Kris 2009 finale vote.

Is the 36 million votes from Siobhan's elimination?

How fast can someone (a tween) text in a vote? I don't know how fast text voting can occur and thus how many, if 1 text at a time, someone could send in 120 minutes. Is the turnaround time faster than a phone vote (using redial of course)? Personally, using a phone I voted 100 times after Sio's AMOM show, in well less than half an hour, and taking breaks. I estimated I probably could have voted 500 times successfully had I pushed through 2 hours. Doing 1-vote texts, how many could a person send?

The answer is useful because it gives us a ballpark estimate of what their cut-off is for limiting votes. "Sandy" assures us that doing single calls or texts you couldn't reach their allowable ceiling. And if you send 5 votes per text for 2 hours, you could be able to reach their allowable ceiling. Could... even then apparently it's not certain.

That strikes me as an awful lot of votes needed to reach their "powertexting" cut-off. Again, can a single-vote texter send more votes than the 500 I believe I could if I tried by phone? Well, 5x that would be over 2,500 votes, which strikes me as a very high ceiling and much higher than what it sounds like you think it is.

Having a very high ceiling also makes sense if they do actually have the policy of looking at each case that hits any ceiling, as opposed to just automatically cutting that number off. I can't imagine the size staff they'd need in a very short time window if you had only just 50,000 or so tweens across the world powertexting like mad. Many thousand would hit the ceilings and need inspection.

Where they put that cut-off line does matter, a lot, because it became no secret that powertexting votes was the way to have a big impact.

36 million votes (if that was the number Sio's elimination night) divided by 6 idolettes is an average of 6 million votes per. The top 2 to 4 competitors will have more than 6 million votes and the bottom 2 to 4 will have fewer than 6 million. The difference between finishing 6th (Siobhan) versus 5th (Casey) places could have been between singers maybe falling in the the range of 3 to 4 million votes each. All guesswork, of course, but AI damns us to this kind of guessing because they won't realease data.

Staying another week versus leaving probably came down to a difference of less than 1 million votes separating Siobhan and Casey, in all likelihood.

If this company's ceiling on votes is, for example, 2500, then just 400 Casey-luvvin' teen-aged girls skilled at powertexting could supply him with a million-vote advantage over Siobhan. Just 400 hormone-charged girls, out of 20 million viewers.

If your average Siobhan voter was casting 10 votes, 400 determined girls-with-cells, at 2,500 votes a pop, could've outvoted 100,000 of them. Leveraged a 250-to-1 difference in individual voting effect.

And from what "Sandy" describes, I wouldn't be surprised if the allowable ceiling before votes start getting discounted is higher than that... it might be 3000 votes. If the company really does what she says and they look, even briefly, at each case of voting past a certain level, they would nearly have to have a ceiling that high.

It matters tremendously what that ceiling is, before power-texting controls kick in. And I'm glad to read the the above communication, as it provides us with better clues on that than I, at least, had ever seen before.

Anyway, as it stands now, I'm concurring with Brian's basic take on this. That is:
(1) A small % of viewers actually vote, at all.
(2) Most of those who do vote, vote only a few times.
(3) Leaving open the ability of a relatively small number of power-texting 2-hour voters to have a hugely outsized effect on the outcome, and making them easily able to decide the contest between last place and next-to-last-place competitors who usually are in the range of 1 million votes apart, or sometimes less.

This is maybe not the ideal thread for this discussion -- we've probably got one somplace -- but it is an important issue and therefore well worth discussing.

So I'm interested in hearing the counter-argument. Pls correct any numbers and counter whatever you think is wrong here.

This post has been edited by blacklightposter: Oct 1st 2010, 7:09 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Oct 1st 2010, 11:37 PM
Post #14


 
****

Group: Members

Posts: 146
Gender: Male



QUOTE(blacklightposter @ Oct 1st 2010, 8:01 PM) *

Jack, a few questions (...then a string of observations):

Aren't those news reports and anecdotal stories from earlier, such as during AI high points? I never followed all of that closely, but didn't votes in a finale or two hit crazy numbers like well over 100 million? I seem to recall something like that during the Adam-Kris 2009 finale vote.

Is the 36 million votes from Siobhan's elimination?

How fast can someone (a tween) text in a vote? I don't know how fast text voting can occur and thus how many, if 1 text at a time, someone could send in 120 minutes. Is the turnaround time faster than a phone vote (using redial of course)? Personally, using a phone I voted 100 times after Sio's AMOM show, in well less than half an hour, and taking breaks. I estimated I probably could have voted 500 times successfully had I pushed through 2 hours. Doing 1-vote texts, how many could a person send?

The answer is useful because it gives us a ballpark estimate of what their cut-off is for limiting votes. "Sandy" assures us that doing single calls or texts you couldn't reach their allowable ceiling. And if you send 5 votes per text for 2 hours, you could be able to reach their allowable ceiling. Could... even then apparently it's not certain.

That strikes me as an awful lot of votes needed to reach their "powertexting" cut-off. Again, can a single-vote texter send more votes than the 500 I believe I could if I tried by phone? Well, 5x that would be over 2,500 votes, which strikes me as a very high ceiling and much higher than what it sounds like you think it is.

Having a very high ceiling also makes sense if they do actually have the policy of looking at each case that hits any ceiling, as opposed to just automatically cutting that number off. I can't imagine the size staff they'd need in a very short time window if you had only just 50,000 or so tweens across the world powertexting like mad. Many thousand would hit the ceilings and need inspection.

Where they put that cut-off line does matter, a lot, because it became no secret that powertexting votes was the way to have a big impact.

36 million votes (if that was the number Sio's elimination night) divided by 6 idolettes is an average of 6 million votes per. The top 2 to 4 competitors will have more than 6 million votes and the bottom 2 to 4 will have fewer than 6 million. The difference between finishing 6th (Siobhan) versus 5th (Casey) places could have been between singers maybe falling in the the range of 3 to 4 million votes each. All guesswork, of course, but AI damns us to this kind of guessing because they won't realease data.

Staying another week versus leaving probably came down to a difference of less than 1 million votes separating Siobhan and Casey, in all likelihood.

If this company's ceiling on votes is, for example, 2500, then just 400 Casey-luvvin' teen-aged girls skilled at powertexting could supply him with a million-vote advantage over Siobhan. Just 400 hormone-charged girls, out of 20 million viewers.

If your average Siobhan voter was casting 10 votes, 400 determined girls-with-cells, at 2,500 votes a pop, could've outvoted 100,000 of them. Leveraged a 250-to-1 difference in individual voting effect.

And from what "Sandy" describes, I wouldn't be surprised if the allowable ceiling before votes start getting discounted is higher than that... it might be 3000 votes. If the company really does what she says and they look, even briefly, at each case of voting past a certain level, they would nearly have to have a ceiling that high.

It matters tremendously what that ceiling is, before power-texting controls kick in. And I'm glad to read the the above communication, as it provides us with better clues on that than I, at least, had ever seen before.

Anyway, as it stands now, I'm concurring with Brian's basic take on this. That is:
(1) A small % of viewers actually vote, at all.
(2) Most of those who do vote, vote only a few times.
(3) Leaving open the ability of a relatively small number of power-texting 2-hour voters to have a hugely outsized effect on the outcome, and making them easily able to decide the contest between last place and next-to-last-place competitors who usually are in the range of 1 million votes apart, or sometimes less.

This is maybe not the ideal thread for this discussion -- we've probably got one somplace -- but it is an important issue and therefore well worth discussing.

So I'm interested in hearing the counter-argument. Pls correct any numbers and counter whatever you think is wrong here.


36M was reported by Ryan on 4/28.

With a GoPhone (right model) texting 3,000 to 5,000 times is doable (how about two or three GoPhones shared with family or friends?)

I voted 860 times using a single text GoPhone.
________

"Vote for the worst with a base of 200K members could have cast 20M votes themselves by voting 100 times."

36Mil is such a puny number for the #1 show on television with a young audience of 20M viewers. I look at all the circumstantial evidence (observations) and see millions of votes lost because of power voting and VoteForTheWorst.

If so few people were voting there would be no busy signals. There would be no reason for AT&T to sponsor a show that didn't provide a R.O.I..

BTW: Sandy may no longer be employed with the VMC and her response was very guarded and vague but revealing none the less...

______


“Portions of this program not affecting the outcome have been edited. The producers in consultation with an independent vote management company, reserve the right to remove so called “Power Dialing Votes” (either toll free or text) that are identified as having been cast in such a significant block, either by technical enhancements or otherwise, that it could unfairly influence the outcome of the voting”

______

http://www.washingtonpost.com/wp-dyn/conte...5052600058.html

Making his "American Idol" finale debut, Edward Boddington, president of the voting management company that administers the show's phone voting system, told the audience that more than 500 million votes were cast Tuesday night for Underwood and Bice.

Assuming the voting was close, that would mean that Bo and Carrie each received about four times as many votes as President Bush did in the last election. Though, of course, you can vote only once for president but you can vote as many times as you like for "American Idol."

______

Still think 36 M was the real number of votes cast?


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
A Song To Sing
post Oct 2nd 2010, 12:04 AM
Post #15



*********

Group: Members

Posts: 2,284
Gender: Male



The Washington Post article was mistaken. I looked up the Season 4 finale on Youtube, and the president of Interscope says that 500 million votes were cast during the entire season. That's far too high a number for one night of voting. By way of comparison, I believe around 100 million votes were cast during the Season 8 finale.

Skip to 1:35...



This post has been edited by A Song To Sing: Oct 2nd 2010, 12:08 AM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
Telement
post Oct 2nd 2010, 7:13 AM
Post #16



**********

Group: Members

Posts: 3,292
Gender: Male
Fav. AI9: Siobhan Magnus



Duh.....sorry, but i dont think AI pick winners based on votes alone. They want a person who is willing to listen and do exactly what they say. Ppl who dont ask too many questions & easy to work with - thats what they want. Kelly Clarkson & the runners up of S2 (forgot his name) sued 19e because they felt they were being cheated by 19. Im sure 19 has learned from all these and will make sure they pick the stupid-dumbass-blur-iwilldoanythingyouwant as a winner whistling.gif

They just use the votes to see who is popular (as a guide for them) and also get millions from AT&T. You think they cant check how many times you vote? They know which number voted 10 times or just once. End of the day, i believe they just take 1 vote from each phone number, becos thats the true representation of votes for the a contestant. So, in future...if you wanna vote, just vote 1 time flowers.gif


--------------------
IPB Image
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Oct 2nd 2010, 3:20 PM
Post #17


 
****

Group: Members

Posts: 146
Gender: Male



QUOTE(Telement @ Oct 2nd 2010, 8:13 AM) *

Duh.....sorry, but i dont think AI pick winners based on votes alone. They want a person who is willing to listen and do exactly what they say. Ppl who dont ask too many questions & easy to work with - thats what they want. Kelly Clarkson & the runners up of S2 (forgot his name) sued 19e because they felt they were being cheated by 19. Im sure 19 has learned from all these and will make sure they pick the stupid-dumbass-blur-iwilldoanythingyouwant as a winner whistling.gif

They just use the votes to see who is popular (as a guide for them) and also get millions from AT&T. You think they cant check how many times you vote? They know which number voted 10 times or just once. End of the day, i believe they just take 1 vote from each phone number, becos thats the true representation of votes for the a contestant. So, in future...if you wanna vote, just vote 1 time flowers.gif


Jaded but true?
Jaded but true!

Maybe that's why... they reported 36M votes on 4/28. It was the number of ESN's & Phone Numbers that voted once.


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
ColonelKlink
post Oct 3rd 2010, 5:14 PM
Post #18



*********

Group: Members

Posts: 2,901
Gender: Male
From: BobbyBobBob Klink



QUOTE(jackxyz @ Oct 1st 2010, 3:24 PM) *

Look... the answer to the question is 36 million votes and 20 million viewers.

So what is the question?

How do 20M viewers cast 36 million votes in 120 minutes?

The answer boils down to certain numbers of potential combinations.

20M ÷ 36M votes = 1.8 votes per viewer

Or

If only 10% of viewers actually vote then 2M ÷ 36M = 18 votes per viewer who took the time to vote.

5% of viewers take the time to vote = 1M ÷ 36M = 36 votes per viewer who took the time to vote.

See a pattern?

2.5% of viewers vote (97.5% don't bother to text or phone in a single vote) =
500,000 voters ÷ 36Million votes = 72 votes per viewer over 120 minutes = 1 vote every 100 seconds.

Even in the worst case scenario where only 2.5% of voters vote over 2 hours that means the average voter stops at 72 votes.

No way is the vote management company acknowledging the votes from people voting many hundreds and sometimes thousands of times.


What do you mean "if only" and "worst case"? I think that you're seriously overestimating viewer involvement in terms of voting. I think it would be miraculous if 2.5% of viewer actually voted, or voted more than once.

Run your numbers with more realistic response rates. Getting 1 person out of 1,000 to respond to information that they requested in the mail is a good yield rate. Next factor in that male viewers of AI are notorious non-voters.

If the voting data was ever released I wouldn't be surprised to see that there were only 100,00-500,000 distinct voters who voted over 20 votes with the remaining voters throwing a few votes in before they got up to fix themselves a snack or prepare for Grey's Anatomy or whatever comes on after AI finishes an episode.


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
jackxyz
post Oct 3rd 2010, 6:56 PM
Post #19


 
****

Group: Members

Posts: 146
Gender: Male



QUOTE(ColonelKlink @ Oct 3rd 2010, 6:14 PM) *

What do you mean "if only" and "worst case"? I think that you're seriously overestimating viewer involvement in terms of voting. I think it would be miraculous if 2.5% of viewer actually voted, or voted more than once.

Run your numbers with more realistic response rates. Getting 1 person out of 1,000 to respond to information that they requested in the mail is a good yield rate. Next factor in that male viewers of AI are notorious non-voters.

If the voting data was ever released I wouldn't be surprised to see that there were only 100,00-500,000 distinct voters who voted over 20 votes with the remaining voters throwing a few votes in before they got up to fix themselves a snack or prepare for Grey's Anatomy or whatever comes on after AI finishes an episode.


You're helping me prove the point.

If AI is getting 20 votes from the average viewer why would they accept my 860 votes or Mac's 3000 or ? ? ?

They want to make sure power voters don't control the outcome. Maybe Tim was eliminated when he was because he had the most power votes about 500 (per ESN & phone #) that week. Unfortunately the following week (with the help of vote4theWorst) Siobhan had the most power votes.





--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
blacklightposter
post Oct 3rd 2010, 9:57 PM
Post #20


 
******

Group: Members

Posts: 543
Gender: Male



To answer Jack's question of why they'd allow powertext votes over a few hundred in number, which is the limit Jack hypothesizes, for the following reasons:
1. It would be too much work to actually do what they say the do, i.e. inspect each case of voting exceeding their ceiling before discounting the votes over the ceiling.
2. It would involve them throwing out huge numbers of votes. The number counted would be much lower than the number discounted, in proportions like what Jack hypothesizes. AI would have a major PR problem if word leaked out that they were throwing out MOST votes cast.
3. To not dampen the passion too many passionate AI fans. Only a small number work hard voting for 2 hours, but those most ardent of fans are important to a show. If they set the powertext vote ceiling low, larger numbers would be frustrated to have their efforts nullified, and they'd quite voting sooner. The show would feel the lessened passion in other areas, and AT&T would feel the loss in cellphone business.
4. To allow AT&T to continue to sell cellphone contracts and especially with unlimited texting options.

To answer the question of why AT&T would stay a sponsor of AI without the 2 hours per week of millions of calls and texts: For the same reason Ford and Coke are sponsors. To advertise their product. People don't gorge on Coke and Ford cars 2 hours every week.

To answer the charge that AI chooses the winner itself and just uses viewer voting for rough guidance, I say, ya'll believe there was a second gunman on the grassy knoll too, don't you? Maybe even that the 9-11 attacks were planned by the U.S. government? Etc. I don't trust wealthy corporations to be honest with me any more than the next person does, but I also don't go in for conspiracy theories. The reason AI doesn't just secretly choose the winner themselves is simple: They couldn't get away with it. Any secret like that eventually leaks out... somebody spills the beans. And then that would be the end of the show. So the reason they don't is that they couldn't do so with impunity. The downside is far, far too steep. And AI isn't Sopranos Sanitation Inc of New Jersey, it's the highest-visibility show on tv. What you're suggesting could not be gotten away with for more than a few weeks before it leaked out. Also, it would be fraud and therefore against federal regulations. FCC would not allow it: even a gutted, post-Bush federal regulator would still leap to investigate and prosecute that massive of a fraud on the public. (Rent Quiz Show to see a 1950s parallel.)

And by the way, as for VoteForTheWorst, it may have totalled over a lot of time, 200k people who registered at some time or other and never un-registered. But that is a far, far cry from actually mobilizing many thousands of "members" to mass vote every week. I do not believe, at all, the suggestions of their voting numbers. Not by a very long shot.


Back to playing with numbers.

Jack texted -- single texts, I'm assuming -- 860 times in 2 hours. (I think I can phone-dial in 500.) Combine this information with what "Sandy" tells us.

She tells us texters like Jack are nowhere near the ceiling cut-off. And that if 5 texts were being sent per message, then someone "could" hit the ceiling in 2 hours. Let's go with 800 texts per 2 hours. 800 x 5 = 4000 powertext votes. You "could" hit the ceiling that way.

So I think my first guess at the ceiling as 2500 votes was too low. I think it's more likely to be 3000 or more.

Again, we all wish AI would release some numbers and let us know factually what?s really happening, but I concur with Brian and Klink. I'm not buying the theory that many powervotes lead to those votes being discarded in huge numbers, and somehow to the recipient of such votes thereby being eliminated. That just does not make sense.

Here's my wild assed guess at, ballpark, the kind of thing that could have happened.

Lots of people watch, but few actually vote, and of those who vote, most vote just a few times.

20 million viewers watched Sio's boot-week performance show.
Fewer than 1 million of them actually picked up a cell or landline to vote.
(I myself watched AI seasons 3-8, with interest, but without voting once. My first and only votes have been for Siobhan.)
Go with 1 million.
Estimate that 90% of those voted an average of only 5 times. That's just 4.5 million votes from 900,000 people.
That leaves 31.5 million counted votes to go that night, to be cast by 100,000 viewers: the more committed ones.
If almost all of those 100,000 people vote an average of 200 times each (i.e. double what I did), that's 20 million more votes. We're up to 24.5 million total. 11.5 million counted votes to go.
If votes exceeding 3000 from one source are discarded, that means that 4000 ultra-committed, hitting-the-ceiling powervoters, from among 1 million voters total, and 20 million viewers, would put you at about 36 million counted votes.

I concur with Jack to the extent that some people had to hit the ceiling, whatever it was, and their votes, in numbers we cannot know (like so much else), were discounted. I.e., that more than 36 million votes obviously came in. But I would not guess the number hitting the ceiling to be too high, for the reasons given earlier -- it would create a problem for both the vote company, and AI, if many millions of votes were discarded.

Maybe, for example, most of the 4000 to hit the ceiling voted an average of 4000 times, so 1000 of their votes each were discounted. That'd be 4 million votes discarded right there. 40 million cast, "only" 36 million counted, and 4 million lost to the powervote controls. Again, all just shots in the dark at guesses, but to illustrate the kinds of levels of voter participation that could be involved, to get to 36 million votes.

You're left with a situation in which relatively small numbers of powervoters can easily decide outcomes, especially between last-place and next-to-last-place contestants who finish not too many votes apart.

It's a broken system. I'm concluding the powervote ceiling and cut-off is high, and I believe Casey had to have benefited from powervoting a lot more than Siobhan did, meaning that it helped to eliminate her.

The voting can be fixed, AI has the power to limited votes per phone. All they need is the management will to do it, and the willingness to have a little less enthusiastic sponsor in AT&T.

This post has been edited by blacklightposter: Oct 3rd 2010, 9:58 PM


--------------------
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

4 Pages V  1 2 3 > » 
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:



-
Lo-Fi Version 81 Time is now: Oct 20th 2014, 5:53 PM




We now accept check/money order donations in addition to online donations via PayPal!  Help us pay for our servers, donate to Neurox Media!  Thanks!
Links: Reality TV Links - American Idol | SirLinksalot: American Idol

Powered by: Invision Power Board v2.1.7 © 2014  IPS, Inc.
Licensed to: Neurox Media