Nate Gowdy

Jenny Durkan is leading Cary Moon 51 percent to 26 percent among likely voters, according to a new poll from Washington State Wire. However, there are reasons to doubt the results; the poll was only conducted on people who responded to landline calls. You can see how this method would leave out a lot of people.

If you believe the results, though, they're bad news for Moon. People who didn't vote for either Durkan or Moon in the primary aren't exactly flocking to Moon, poll suggests.

"Some Seattle political observers had thought that the non-Durkan voters in the primary would likely move towards Moon as a more progressive alternative in the general election. That does not appear to be happening," the poll says.

The results also show that former mayor Ed Murray's resignation after allegations of child sex abuse didn't matter much to voters in making their decision — and in particular, that if Moon is trying to tie her opponent to the former mayor, that strategy isn't working. Only 15 percent in the poll said Murray's resignation mattered for their vote, and of these, about the same amount supported either Durkan or Moon.

Washington State Wire's findings contrast with a poll conducted by the Moon campaign earlier this month that showed the two candidates neck-and-neck.

The survey was done between last Saturday the 23rd and Tuesday the 26th and is based on 510 results, with a 4.33 percent margin of error and a 95 percent confidence rate. You can read more on the methodology here.

There are other reasons to take the results with a grain of salt. A poll from Washington State Wire in June overestimated the support for Bob Hasegawa and Mike McGinn, who finished fifth and sixth in the primary, respectively (though it should be said that poll did find a large percentage of uncertain voters).

The reason landlines were used is that it's illegal to call cellphones with automated messages (thankfully) — but still, the results probably suffer for it.

As HuffPo pointed out last year, 64 percent of the population uses cellphones rather than landlines most or all the time, and among young people aged 25 to 29, more than two thirds lived in a place with no landline. White people are most likely to live in a house with a landline phone, and among Latinos, 60 percent lived in a cellphone-only house.

A New York Times report notes the rise in cellphones and decline in people answering surveys as challenges, but another is that even if people say they plan to vote, they might not.

Washington State Wire addressed the issues with landline polling on Twitter, saying the concerns are "overblown." The group noted that most voters are older to begin with.

Still, the group acknowledges that this survey is not meant to be definitive. As the methodology section of the poll notes: "Given the complex nature of determining public opinion in these races, this survey should be used as only one data point among others to determine with high confidence the direction of Seattle voters ahead of the November general election."