Since we have been all polls all the time why change today? Wall Street Journal/NBC/Marist has arrived with three key swing state numbers, from New Hampshire, Iowa, and Wisconsin. They all show President Obama holding onto small leads that have diminished over the past month. The numbers please.

For me I look to Wisconsin first. Even though some of my friends don’t believe it I still concede that Mitt Romney has a chance to win this election. I try to look at the numbers without trying to spin them so that I get a result I am hoping for. Not to be a broken record but I don’t see how Mitt Romney wins without Ohio. That is not a partisan observation. I think Mitt Romney believes it too. Wisconsin is important because I believe that if Romney were to break through here it might indicate that the Obama mid-western firewall, including Ohio, might be in trouble. So lets see what Marist is showing in WI.

25% of voters have already voted in Wisconsin, and as with other early vote states the President is leading by large margins with those voters. (59% to 39%). Overall the President leads Mitt Romney, with likely voters, by a 49% to 46% margin, with 3% undecided, and 2% with “other”. That 2% looks pretty important to both candidacies right about now. It is also clear that voters are hardened, and there is not a lot of persuasion left to be done. The number is within the margin of error, and shows Romney cutting the President’s lead in half since the last survey. An interesting note is the favorable/unfavorable number. The President stands at 53%/44%, while Romney is at 47%/47%. The President’s job approval is at 49% approval, 45% disapproval. And the President has a big lead with women, leading here by 14% with females. The big US Senate race in WI has Democrat Tammy Baldwin with an ever so slight lead on Tommy Thompson 48% to 47%, with 1% undecided (talk about hardening) and 4% with “other”. That 4% looks pretty big right now as well. Another key race that may well determine the majority in the US Senate.

In small but critical New Hampshire Obama maintains a slight lead, which is also down from the last Marist survey, and within the polling margin of error. Obama leads 49% to 47%, with 3% undecided, and 1% with “other”. The President’s job approval is at 48%/48%, while his favorable/unfavorable is at 50%/46%, while Romney is at 49%/46%. One other note is the NH right track/wrong track question, which is at 42% right track, 53% wrong track. That is a negative for Barack Obama in light of the tightness of this race. In NH the President leads with women by a whopping 16%. For all the talk of closing the gender gap it appears that if Barack Obama does win re-election then one of the major pillars of that effort would be his support with women. New Hampshire is real close, and remains a true toss-up state.

The New Hampshire Governor’s race shows Democrat Maggie Hassan beginning to open up a little distance between herself and Ovide Lamontagne, the Republican. Hassan leads with likely voters by a 49% to 44% margin, with 6% undecided. A Hassan victory here appears likely.

The survey shows that 10% of the respondents in NH had voted early, and with that group the President leads by a 56% to 42% margin.

I had moved NH from toss-up to Obama on my last map, but you can see it will be a very tight race there. I also have Wisconsin in the Obama column, but I feel more confident there. I did not talk about the third leg of this story, Iowa, because I see that as being a bit out of reach for Romney.

Even with the very small margins I am sure the Romney folks will call into question the “methodology” of these polls. Democrats over-sampled, likely voter screen too strong/weak, pollster’s mother is a Democrat, and on and on we go. In light of all of these objections I have searched for a poll that would be considered “properly sampled” by Republicans, and I believe I have found it. It is the survey that shows Mitt Romney in the lead. Unfortunately that survey is not from Ohio.