Do you really think that there will be more than 57% turnout for a mid term election? Are you insane, or just silly?
Talking about a poll not turnout and only 2 months out you'd think they'd only be polling likely voters?
Likely voters are used as a measure to determine what most likely will happen in the final votes is far more reliable on a statistical basis than just anyone you find in a phone book. That is why the most reliable pollsters like Rasmussen uses likely voters and then weights that to the statisitcal makeup of the areas polled to arrive at a reasonably accurate prognostication.
It is also a statistical given for elections, that a result for just anyone will be far more progressive leaning than one comprised of likely voters.
So in these Survey USA results that polled 1000 adults, 844 were registered to vote so right away the results are questionable. Of all of them only 569 were likely voters so again the error potentional was again increased.
A typical liberal poll will us the results from the full count for PR purposes because it is known that not using only likely voters will skew a poll to the liberal side.
So your response was not appropriate to say that 57% would show up. That was not the implied point. The point was this poll is most likely wildly innacurate with weight given to the liberal candidate, which, makes the chances of a R win more likely.