These statewide polls with their statistical models and methodology are not reflecting the huge shift toward Romney in Ohio coal company. There is a turn around from 2008 of more than 50,000 votes. This is very significant is a state that may be decided by two or three thousand votes overall.
Are reflecting or are not reflecting?