Bear in mind that Trump remains popular there and will campaign aggressively for Moore this fall if need be, knowing how catastrophic it would be to fumble away a Senate seat in the deep south when McConnell’s margin is already razor-thin. A Democratic upset here would build liberal morale for the midterms like nothing else. Trump will do whatever Moore needs to prevent it from happening.

Also bear in mind that the Democratic nominee, Doug Jones, is a stone-cold abortion fanatic, seemingly willing to protect the right to choose right until the moment of birth. Comparatively few Alabamans know that about him right now, I’d bet. Many, many more will know it before polls open.

Having said all that, though, Moore leading Jones 50.2/44.5 is closer than you’d ever expect this race to be:

Moore boasts a huge lead among evangelical voters, 67.8% to 28%, while Jones boasts an even larger lead among non-evangelicals, 69.7% to 26.7%. Among African-American voters, Moore peels off 24.8% to Jones’ 70.9%, while among white voters, Jones has a surprising 36.1% to Moore’s 58.5%…

Moore underperformed Republican Presidential candidate Mitt Romney by eighteen points, losing Madison and Mobile counties and winning on the strengths of Republican rural pockets. It is entirely possible that he has a closer contest than anticipated this year as well: it’s a special election, it’s in December, and Democrats have overperformed in every Congressional special election this year.

But, it is Alabama, and Moore has a dedicated base of evangelical voters that carried him through two rounds of primaries and won’t abandon him over the next eleven weeks.

An alarming number there for both sides is Moore pulling nearly 25 percent of the black vote, a surprisingly large share for a Republican. It’s hard to measure that against other Republicans in Alabama since exit polls of the state aren’t easy to come by. Usually there’s no point taking them since electoral results there are rarely competitive. (Jeff Sessions ran unopposed in 2014.) It may be that Moore’s reputation as an unusually stalwart social conservative, a man willing to defy a federal court to keep the Ten Commandments on public land, has earned him a larger-than-usual following among black voters for a Republican. If so, Jones’s task is even more difficult than he knew. The risk for Moore, obviously, is that if his black support erodes and starts to shift to Jones, the overall margin is already close enough that he’d be in deep trouble. Jones is no random Democrat on racial issues either: He prosecuted the Klansmen behind the 1963 Birmingham church bombing. If he takes, say, 85 percent of the black vote and manages to hold onto the 36 percent of whites he currently enjoys in this poll, the GOP will be at risk of a momentous upset.

The dirty little secret about Moore, as noted above, is that he’s not a particularly strong candidate even in Alabama. Conventional wisdom has it that his anti-gay, anti-Muslim rhetoric makes him hyper-popular in the deep south, all but unbeatable amid the GOP’s populist fever, but that’s just not true. Even in ‘Bama he’s a gamble. In addition to grossly underperforming Romney in winning back his court seat in 2012, he ran for governor against a Republican incumbent in 2006 hoping to capitalize on his Ten Commandments fame. He got crushed in the primary, losing by more than 30 points. He ran again in the gubernatorial primary in 2010, among a field of four Republican candidates. He finished fourth. To the extent that he’s unbeatable in December — and he probably is — it’s because this is an important race and he’s a Republican in an overwhelmingly Republican state. It’s hard for the GOP to lose Alabama under any circumstances. In fact, Moore by six may realistically be close to the worst-case scenario:

Moore’s been ripped to shreds on Alabama TV over the past three months while no one’s laid a finger on Jones yet — and Moore’s still north of 50 percent, even here. That’s a heavy lift for Democrats.

It occurs to me that Steve Bannon’s gamble on Moore against Luther Strange was *almost* totally risk-free. Bannon jumped in for Moore late, when he was already comfortably ahead of Strange in runoff polls. There was every expectation that he would win. By endorsing him loudly, Bannon was able to exploit Moore’s victory as a victory for populism writ large, which ideally would scare Trump back towards the right and would encourage populist candidates to primary Republican incumbents in their own states. If, on the other hand, Moore had ended up losing after Trump intervened in the race, that would have been no great setback for Bannon. He would have pointed to it as proof that Trump, the populist-in-chief, still commands supreme loyalty among Republicans. Populism won because Trump won again (even though his guy, Strange, was no populist). Bannon’s wager was essentially “heads I win, tails you lose,” at least for spin purposes.

There is a way he could lose big, though. If Moore ends up being upset by Jones, it’ll ratify every establishment warning that the party is nuts to bet on loose cannons in primaries against mainstream incumbents. Moore will be compared to Todd Akin, Christine O’Donnell, and Sharron Angle, grassroots favorites who lost otherwise winnable races. McConnell will solemnly intone that Big Luther would have won, which is probably true; a bland generic Republican probably beats Jones easily in Alabama for purely partisan reasons. Bannon will find someone in the Republican establishment to scapegoat if Moore fumbles the race in order to reassure Republican voters that populism remains the path to ultimate victory, but not everyone will buy it. I guarantee you there’ll be stories here and there before December of Republican operatives quietly assisting Jones’s campaign behind the scenes, hoping to keep a headache like Moore out of the Senate and to deal Bannon’s populism crusade a heavy blow before it gets rolling. But no one’s going to take that risk unless there’s reason to believe Jones can win. There really isn’t. Yet.