are different than some of the national public polls, that look at turnout scores and turnout modelling. but that's based on past voter propensity. you talk these estimates into account and look at the different range of turnout scenarios to see what the range of outcomes will look like. but -- and then we have now early voting which in some states is starting to look democratic, and some states it's actually a little more republican. is that a signer where things end up? or is that just election day voters who are just voting early as opposed to election day. these are all questions we won't know the answer to yet. so that is oop why it's important to not just look at who is up plus 2 or what does the model say about this, but what are the issues that are driving voters? what are they thinking about? what calculations are they making? enthusiasm is one metric, but an enthusiastic voter's vote counts the same as someone who reluctantly goes the day before