According to a new MPR/Humphrey Institute poll, the race is now deadlocked. To some extent, the race was always going to tighten up once all of the candidates firmed up. On the other hand, a significant amount of the change can be attributed to the poll’s methodology.

Here are the poll’s results, for what they’re worth:

Mark Dayton: 34%
Tom Emmer: 34%
Tom Horner: 13%
Undecided: 19%

Now, what are the issues with the poll? First of all, it was a somewhat small sample of 750 people, with an estimated margin of error of 5.3 percentage points. That means one candidate could have as much as a 10-point lead, even if that is fairly unlikely. Even worse is the methodology: The poll was based on a “landline, random-digit dial survey.”

Landline? Are you kidding me? I wonder how many younger voters were missed. Not having a landline, I could never have been contacted for this poll.

Maybe that helps to account for the poll’s likely-voter model:

Republican: 46%
Democrat: 41%
Independent: 13%

You’ll have to excuse me if I don’t buy that. The oversampling of Republicans is yet another reason to suspect that this poll overstates the extent to which the race has actually narrowed.

Ultimately, this is just a single poll, and time will tell whether it’s an outlier. I suspect its broad finding — the race narrowing — is probably accurate. But I also suspect that future polls will continue to show a Dayton lead, just a somewhat smaller one.