Ask me why I drink.

I’ve decided that this is the only race I care about on Tuesday. The incoherent polling has bred in me an irrational emotional investment in the outcome. New from Vox Populi:

Sinema by four? That’s quite plausible. She led by that exact amount in CNN’s poll of Arizona a few days ago and led by three in a CBS poll taken several days earlier. Both the CNN survey and one from NBC have her at 50 percent or better. A 52/48 Sinema win is very much on the table.

New from Harris Interactive and Scott Rasmussen:

That’s confusing at first glance. The first three columns are the numbers over the last three days, the second three columns are the numbers over the last five days. “BASE” is their best guess at the outcome given the expected levels of Democratic and Republican turnout. “HIGHD” is what the outcome might look like if Democratic turnout is higher than expected, “HIGHR” is what it might look like if Republican turnout is higher than expected. In other words, the pollsters’ best guess is that McSally is up 6.7 points right now, which is a bit tighter than the five-day average. But even in the realistic worst-case scenario of Dems turning out in force, the pollsters still see McSally winning narrowly.

Is that plausible? Well, the last two polls of Arizona taken by another outfit, OH Predictive Insights, had McSally up six and seven points, respectively, right in line with the Harris/Rasmussen numbers. So yeah, this result is on the table too. Not only that, but the early-voting tea leaves look good so far:

Geraghty elaborated on that in a post:

What we don’t know in this equation is how many Arizonans will vote on Election Day. If the turnout is low like in 2014, with about 1.5 million votes cast, Sinema would need to win the Election Day vote by an overwhelming margin, something like a 73 percent to 27 percent split. But if Arizona’s statewide turnout is higher, like in 2010, when about 1.75 million people voted, Sinema could attempt to make up that 112,642-vote margin out of another half-million voters. That would require a 62-38 split in Sinema’s favor — less difficult, but still not easy. An NBC News poll found Sinema leading among self-identified independents, 58 percent to 32 percent.

We always have to be careful about assuming too much from early votes. All we know is the party registration of the voter, not how they voted. An unusual number of GOP crossover ballots for Sinema plus wave-like Democratic turnout on Tuesday might be enough to tip the race blue. But obviously you’re happier with those early numbers if you’re McSally than if you’re Sinema. Four days away: Anything can happen.