American Idol – What Went Wrong?
We were riding an unbeaten streak here at Alchemy. Based on a request that came in from a prospect we decided to take a look at the American Idol match-up. Rarely, do we ever see a model that was so strongly in favor of one party – in this case Bowersox. From that model we felt very confident that barring a mistake in the next to final episode, Bowersox would be crowned the victor. Judging from media coverage, Bowersox had a strong performance the night before the final. In fact, the SoV measure in between the last performance and the finale didn’t change – she still had commanding lead in share – even for that night.
So what happened? We could try to cast dispersions on Idol’s methodology as some have done. But in reality, I don’t think you have to look very far back to learn one thing – the underdog often wins on Idol. Idol voting is not the same thing as voting in an election. You can vote early and often. There is no one viewer – one vote restriction. Thus, fans of the perceived underdog have the ability to outvote fans of the favorite. Going into the final vote, underdog fans know this and are more motivated to place multiple votes for their performer while fans of the favorite are less likely to vote multiple times. Or so the theory goes…
In the end, SoV measures for Idol (and shows like Idol) might be contra-predictive. In fact, using the same SoV measure, Adam Lambert should have won last season. Perhaps, the model going forward is to pick the underdog. Or, perhaps, buzz (SoV) doesn’t equate to votes on American Idol.
Either way, we’ll now turn our attention back to predicting politics.
Tags: American Idol
