I asked Owens spokesman John Boughtin whether they were confident in his win. "It really isn't something that we're looking closely at," said Boughtin, seemingly unconcerned. "He's got a big job ahead of him and that's what we're focusing on."
The issue here is that in any election, initial numbers from precincts can often contain mistakes, due to simple human error by exhausted election workers. As a hypothetical example, a precinct might have 410 votes for a candidate, which could be mistakenly written down as 140 and not proofread until a week later. These errors are distributed at random and thus usually cancel each other out, in terms of the horse race -- but in a very close election, they can sometimes make the difference.
The thing to remember, however, is that it remains highly unlikely that Hoffman could have won, as he would need an extraordinary margin among a small pool of absentee ballots. Only about 10,000 absentees were distributed in total, and some of them will not have been returned on time. Some of the returned ballots will contain essentially wasted votes for Republican Dede Scozzafava, who dropped out of the race the weekend before the election and endorsed Owens, shrinking the pool even further. And of course the Democrats usually maintain active absentee ballot operations, so Owens will have a significant number of votes.
Even if Hoffman were to lead Owens among absentees, a question we simply don't know the answer to, the odds do not speak in favor of any sort of landslide win necessary to give him the race. So Owens can rest assured that in all likelihood, he won this race -- just by a slightly narrower margin than we first thought.