In terms of false getting altered results because of different regions getting saturated with callers, I don’t think this is going to cause much of a problem because the data being collected here is all from hardline calls. There is no cell saturation problem as a result, and while Idol is extremely popular, I don’t think it is going to be able to saturate the wired telephone network. Especially the modern network that does not really utalize circuits in the same way as the old network meaning essentially there is no need for one physical line connecting you and the idol as would have been the case a long time ago. Instead voice traffic is going to be converted to a digital signal pretty shortly after it leaves your home. At that point it is essentially pacaktized data flowing over a data network and so the saturation problems really will be mostly endpoint things and are not going to be overly influenced by source network.

The second thing is the lack of text message voting data, this is a factor but really should not be one that matters over-much because, in theory at least, there should not be too big of a difference between text voters and telephone voters.

On a side note I think this system could be used to produce really acurate results by picking the data utalized more judiciously. As your pointed out above, data gets squed for a variety of reasons, a lot of them having to do with saturation over time. I think it would be possible to get a much greater degree of acuracy if instead of simply utalizing a mass of data, a specific set of data from certain points throughout the dialing period was selected. say 50 calls per contestant at 30 minute intervals. This would provide a smaller, but probably statistically more acurate sample set

Anyway just rambling now so I am done, but I would be interested in seeing if the source site could post their raw data over time and see if that could result in a more acurate sample.

Elliott being at the bottom isn’t a great sign (I’d be disappointed if he goes this soon also), but look at DialIdol’s results from last week. The bottom three were Lisa, Ace, and Kevin, all of whom are still around.

Elliott is pretty alone (percentage-wise) at the bottom this week, but it’s definitely not a sure thing. But don’t come after me throwing Carrie Underwood CDs if he does go tonight…

There might be an argument for certain time zones having fewer viewers and fewer callers (the flyover states’ time zones would have fewer people) and so contestants whose fan base is in those time zones may have an easier time getting through.

In that case, yeah, something else to throw off the results.

I still think it’s a good thing to measure, and it’s been very reliable so far, but there sure are a lot of factors stacked up against it being that accurate.

Of course, maybe it’s a moot point now that DialIdol has been served with a cease and desist letter.