First real earthquake poll that I can recall at the state level, not unlike how Monday’s Pew national poll felt. I’m tempted to break out the mushroom cloud thumbnail, but … no, not yet. Let’s save that for the last two weeks.

Second look at that Suffolk pollster’s decision to stop polling Florida?

The survey conducted this week found 51 percent of likely Florida voters supporting Romney, 44 percent backing Obama and 4 percent undecided. That’s a major shift from a month ago when the same poll showed Obama leading 48 percent to 47 percent — and a direct result of what Obama himself called a “bad night” at the first debate.

The debate prompted 5 percent of previously undecided voters and 2 percent of Obama backers to move to Romney. Another 2 percent of Obama supporters said they are now undecided because of the debate…

Across the board, from who is better suited to improve the economy, to who will protect Medicare, to looking out for the middle-class, to handling foreign policy, likely Florida voters now favor the former Massachusetts governor over the president…

Especially ominous were the numbers for Hispanic voters, a demographic where the Obama campaign is banking on an advantage of at least 15 percentage points. The poll showed 44 percent of likely Hispanic voters favoring Obama and 46 for Romney, though the margin of error is higher with that smaller group of voters.

Among independents, an 11-point lead for Obama last month is now a 13-point lead for Romney, which jibes with the Mitt-mentum you’re seeing within that group across multiple polls. How bad is it for O? Dude:

Whose plans are more likely to do more long-term harm to Medicare? Obama 54 percent, Romney 40 percent.

Can’t tell from the sparse data here whether that result is a byproduct of the state shifting for Romney generally or something more particular to the specific issue as the Romney/Ryan message on entitlements penetrates, but it’s encouraging. I recommend following the link up top and reading the methodology on how the poll was conducted, as Team O is already spinning it as defective. Mason-Dixon, the pollster, uses a sample based on actual registration data to try to match it to the electorate; other pollsters typically don’t, which may help explain why only one other poll has shown a lead this big for either candidate in Florida since early August.

In lieu of the traditional HA celebration cameo from Humpbot, I offer you this new clip via Katie Pavlich and Townhall, which is far more enjoyable.