General Discussion
In reply to the discussion: Nate Silver has some competition [View all]ProfessorGAC
(65,010 posts)Big deal! He found a correlation between someone getting a lot of votes in a primary getting lots of votes in the EC.
Don't need a model, or even math to find that correlation.
But, this does, not at all, establish causation.
The factors that lead to a big primary total aren't necessarily the same that lead to an EC majority. This is a general consensus amount political experts. It's why candidates typically run to the middle for the EC.
It's contingent upon him to demonstrate a causative link, which is not done in the model.
And, for all know, he weighted some thing to assure that 24/26. After all, he claimed to accurately predict elections from over 90 years ago where the result was a known. Easy to tweak a model to get the correct prediction when you already know the result.
There's no valid math behind this, and it's a sloppy & simplistic model.