Romney takes the lead in the Politico tracking poll for the first time since last May.

With 49 to Obama's 47%, Romney has picked up 3 points since the last poll, a week ago.

Women propelled Romney’s move into first place in the poll — a majority of which was conducted before the Hofstra debate. Obama’s 11-point advantage a week ago among the crucially important group dwindled to 6 points. The Democratic incumbent still leads 51 to 45 percent with women, but Romney leads by 10 points among men.

Here's Nate Silver's analysis of the gender gap — "'Gender Gap' Near Historic Highs" — published yesterday, using, among other polls — last week's Politico numbers:

The biggest gender gap to date in the exit polls came in 2000, when Al Gore won by 11 points among women, but George W. Bush won by 9 points among men — a 20-point difference. 

Silver had this year's gender gap at 18, with Politico contributing the number 20 to his average of 9 polls. He depicts the gap with 2 Electoral College maps, "recalibrated" to add 9 points to the polls in every state first to Obama and then to Romney, which might represent a hypothetical in which only women vote and a hypothetical in which only men vote, even though Silver doesn't say that — either because it's upsetting to talk about radical disenfranchisement or because the hypothetical would change the country so radically that it would make zero sense to plug in the numbers from existing polls.

In scenario #1 Obama is more ahead than Romney is in scenario #2. Silver's prestidigitations somehow, once again, hearten the Obamaphiles in the NYT readership, but the underlying numbers have shifted. Why, after all these efforts at making women feel especially bonded to Democrats, has the greater gap developed with men going Republican? Perhaps that question contains its own answer: The blatant appeals to women don't win women so much as they lose men.