Idol semifinals: No way to run an election

Last night on American Idol Elliott got the boot (alas).

They handled the results differently than usual -- they actually announced the percentage shares of the vote for the three contestants: 33.06, 33.26, 33.68. (They didn't identify which person got which of the other two percentages.)

33-33-33 OMG! Cue oohs and aahs from the AI theatre audience at the incredible closeness of the vote. But it's misleading.

It's well known that a huge number of votes gets blocked because even though they provide multiple phone numbers to call for each contestant, all the numbers get jammed for the whole two hours of the voting period at this stage of the contest, so people face busy signals and have to keep dialing to make any votes count. (I kept calling to vote for Elliott and actually got through only 3 times.)

So it appears that the percentages of the vote total are just indicating that the carrying capacity of the phone lines was reached for all three contestants, and what difference there was in voter preference has been artifically collapsed down to look very small because we (and they) are unaware of a massive number of intended votes that are dropped as busy signals.

Put another way, the show's voting infrastructure is woefully insufficient to the load.

On some boards, people have even speculated that the text-messaging votes are what made the most difference because of the flooding of each contestant's phone inputs. (We don't use Cingular, so the text messaging option is out for us.)

I doubt the outcome would be different if all the votes had gotten through; but to present the raw percentages as though every vote was actually counted and shockingly the result was so evenly divided is misleading.

  Music   3 comment(s)  

I think the numbers for the vote are faked. The real winner is already chosen and agreed to demand less income than the other candidates.
      ...posted by Maria loves pictures on May 26, 2006 11:55 AM

Hi there!
Long time no cross-blog!
You still have me on your blogroll... I really must reconstruct mine...

Back on topic, I adressed the American Idol voting capacity issue (among a bunch of other things) in a piece I got weblished two years ago:
"American democracy meets American Idolatry"
(In 2003,) "the big vote and the close vote were dangerously related: the flood of calls overwhelmed Idol's phone system, and with both phone numbers running at capacity, a near-tie was inevitable. The winner seemed to have been decided by the relatively few voters who used AT&T's wireless text-messaging system, which never had capacity problems. Or a technical difficulty lasting less than a minute to Clay's number could have accounted for the half-percent margin. Nobody outside the show knows for sure, and nobody inside the show is talking."
      ...posted by wendell on May 28, 2006 11:46 PM

Wendell, greetings!

Here I thought I was being smart for figuring out the voting-closeness on my own, and you had it pegged three years ago...

Randy said on the Larry King followup show that the finale's vote margin was not so close, which I assume is because they kept the voting open for four hours instead of two - more time for a difference in fan ardor to show.

Good to hear from you. :)
      ...posted by Steve B. on May 29, 2006 12:29 PM

Add a comment...