A bit of X Factor learning, or “Paije against the machine”

X-factor-logo1

Having just blogged about the benefits of constant learning over expertise, I’m happy to declare that I’m still learning…

We watched the X Factor results excitedly, mainly because we were waiting to see who was right – me or the predict-o-graphs. Here’s what I wrote on Saturday:

Matt and Cher still top for me.
Rebecca and Paije – growers
Mary, Treyc, One Direction – solid, probably safe for now
Wagner and Katie – now both on planets of their own where our laws of physics don’t apply, anything could happen
Aiden and Belle Amie – DANGER! DANGER!!!

while computer said Mary had little interest and was trending down, with Paije and Treyc also worryingly quiet on the ratings front. Belle Amie and Katie were next lowest.

So if I’d combined those two into a prediction of who would face judgement (which I didn’t) it would been like this:

Computer says Mary – we say NO
Computer says Paije – we say NO
Computer says Treyc – we say NO
Computer says Belle Amie – WE AGREE!
Computer says Katie – we say “NO IDEA!”, could be…

And it turns out that it was Belle Amie and Katie in those slots, in that order.

Actually, I pick music for radio stations in a fairly similar way – we do some number crunching and some gut feel judgement and bring the two together. So “yay” for that process.

I had another look at the numbers to see if they, on their own, could have predicted that bottom two. I tried separating out the “love” comments from the “hate” comments, but Mary has the least love on iTunes as well as the least ratings overall. I might try a slightly more sophisticated model than counting the reviews, but I’m guessing we will have to conclude that the iTunes crowd doesn’t represent the whole. But we’ll try again next week and maybe be a bit bolder with the predictions – it’s fun! Well, we think so.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>