7Although to be honest, don't forget there is an independent candidate who will likely get 7% in the end.
So a poll like: 45% Conway 40% Bevin 7% Curtis becomes if we only count Conway and Bewin: Conway: 48.5% Bevin: 43.5%. Much closer to the theorical 50% of a dual.
I think what will happen is Curtis will get less than the polls suggest (as usual for third party candidates). He will instead get 4-5%. That leaves 95-96% of the vote for the two of them. If Conway has 45% of it, and Bevin has 40% of it, Bevin could win the election with 48% of the vote (80% of the undecideds, hypothetically). That's being pretty generous though, but after 2014 I've pretty much declared war on polls. Like I said, i'll take into account a lot of things and make a final decision Monday.
I read one poll that said, for example, that Bevin was only leading southern Kentucky by 2 points. Even in a really terrible situation, he should be getting at least 55% in that region with undecideds. So its that kind of fish stuff which leads me to believe the polls aren't very reliable. But then there's big money, ads, that has a Conway advantage (Kentucky is a great state for TV ads, retail politics, etc.) McConnell would've won on big money alone even if he massively screwed up his 2014 race. So there's a lot to think about here.