Polling data in Washington state is hardly an exact science, nor is it done with any kind of regularity.

A recent Crosscut/Elway poll showed Gov. Inslee would get 46% of the vote in an election today, with GOP contender Loren Culp 14%, followed by Dr. Raul Garcia (Tri-Cities ER physician) at 6%, then Joshua Freed (former Bothel Mayor) at 5, and Tim Eyman at 4 percent.

The poll said it utilized 402 registered voters, but 80% were on the West side, 30% of those in King County. Only 20% came from the East Side. We also don't know what parts of Eastern WA were sampled.

Who is Crosscut?  Who is Elway?

Crosscut is an independent news organization, similar to Public TV and NPR, except they're an online news source. They have affiliations with KCTS-9 Public TV, and they're located in Seattle. They do a decent job of media coverage, but do have a decidedly left leaning slant. They recently featured an article that claimed many Seattle area parents believe it's white parents "responsibility" to integrate Seattle schools. They also feature as part of their content sections (besides news, education, culture etc) a category called "equity."

Elway is a polling media group who began operations in 1992. In 2018, they partnered with Crosscut to create their polling service. While it's admirable that they are trying to fill this all too vacant gap, recent work has shown their data is likely just too small of sample sizes to base important political and voting decisions upon. Elway is an outgrowth of efforts that began affiliated with KCTS-9 TV.

By contrast, Rasmussen and Gallup polls utilize far larger sample sizes. Gallup is known worldwide, and Rasmussen is probably considered the most 'believable' of the polling services.

It's difficult to accept that Gov. Inslee has statewide support to the degree of 46% when a May 15 KING-5 TV poll showed his approval numbers were only 41% approval, and 48% of the respondents who identified as independents "disapproved" of his performance.

The point here is, polling is an inexact science, and without the proper sample size, demographic and other factors, can produce some very interesting results.