This is a click-bait headline, as the conclusion of the article is "we'll compare the two prediction methods tomorrow and find out which has done better".

I'll go so far to say that the article itself is lacking content, and does little more to say that two popular models don't agree on percentage probabilities, even though they both agree who has the advantage. Apparently that they both agree on Clinton having an advantage is not sufficient for the author.