CA-52: DeMaio leads by 1 in poll but will win

A friend asked me what kind of VBM number Scott Peters needed to win the CA-52 election. I replied that I don't think there is one that can get him there. This poll has DeMaio by 1 and we'd assume that puts Peters a decent bet to win.

If we look closer, however, we see that DeMaio leads by 12% with those already voted. People who've already voted are 42%R/33%D. That's a very positive sign for DeMaio, because it shows that he's getting the Republican support that some people thought he might lose due to NOM endorsing his opponent and the allegations.

Peters leads by 6% with those who haven't voted and that the final turnout will be 35%D/31%R. I don't see how that's possible. In 2012, the VBM was R+6% and the final vote was R+4%. Here the VBM is more Republican and the final vote skews Democratic. To get that, we'd need a turnout of around 217,000 votes. Based on VBM returns being so low, I'm anticipating !70-200,000 votes. Then we'd need those votes to be 36%D/24%R. Even when Democrats were turning out massively on election day you couldn't get that kind of Republican turnout here. As I've noted, the Democrats started pushing VBM this year and we didn't see the vast difference in VBM/election day voting in either the San Diego mayoral election or in the primary. Republicans got 57.7% of the VBM vote and 57.8% of the election day vote in the CA-52 primary. So the idea that election day would be this Democratic is absurd.

Even with all that, this poll has him losing by 1%. Peters' ceiling, if things break in a way that I don't see possible, is a draw. Based on the VBM, low turnout, and how this poll says that the early vote has gone I don't see how Carl DeMaio loses. I'm going to make predictions in the races but this one is my only take it to the bank certain one.