Top 7 redux … predictions?

I’m just going to throw out there that I don’t really know what the hell is going on.

Although I had a feeling that Jessica could be Pia-ed last week, that’s really all it was. There was no evidence for it. That it came to pass is a testament to the seeming randomness of AI Season 11.

Previously reliable statistical factors like WNTS approval rating are proving to be very bad. Indeed, the numbers today don’t feel right at all. They put Joshua with an average of 48.5 approval (below average), including rating his “A Change Is Gonna Come” at a mere 59. Colton’s numbers were also surprisingly low at 35.5 for the evening; although I don’t personally disagree, I would have assumed a better response.

On the other hand, Dialidol has shown some slightly better than average picks, though not amazing in any way.

For the past two weeks Hollie was the call. This week Hollie is predicted to be safe, so she’s probably going to get eliminated.

The combination of statistical factors being what it is, the model assigns a significant probability to all contestants. All 5 of the people who’ve been in the bottom 3 have basically the same probability of being eliminated. Colton has a lower, but significant 9% chance of being eliminated. Phil at 5% feels a little low, but not ridiculous.

My personal guess is Elise.

I don’t regard Jessica Sanchez as safer than any other contestant, by the way. The people who have been saved have never been eliminated right away, but reasoning based on 3 data points is not sound. This is especially true when considering how early 2/3 of those saves were (Michael Lynch in the Top 9 and Casey in the Top 11). Moreover, Matt Giraud was saved in the Top 7, and while he did survive the Top 7 redux, he was then eliminated in the very next round.