Reuters/Ipsos Natl: Trump +2

[Seriously?]

Note that this is LV, previously Reuters/Ipsos would report out their RV number, but I guess after the conventions they switch to LV. Last week's reported poll (RV) was Clinton +4.

Also from the article (assuming from their LV model)...

You must be logged in to read this quote.

They don't have polling explorer loaded yet, but LV was trending to the left of RV last week, IIRC, it flipped during the convention.

Trump was down by 15 amongst LV on July 14 in this poll. So Reuters came back down to Earth very quickly.

[Seriously?]

This poll continues to not make much sense.

2-way Likely Voters
Trump 40.2%
Clinton 38.5%
Neither/Other 13.1%
Refused 6.9%
Wouldn't Vote 1.2%

2-way Registered Voters
Trump 38.3%
Clinton 37.6%
Neither/Other 14.4%
Refused 7.4%
Wouldn't Vote 2.3%

4-way Likely Voters
Clinton 39.9%
Trump 37.0%
Johnson 6.7%
Stein 3.1%
Neither/Other 4.1%
Refused 7.2%
Wouldn't Vote 1.9%

4-way Registered Voters
Clinton 37.6%
Trump 36.2%
Johnson 7.2%
Stein 3.2%
Neither/Other 4.8%
Refused 8.0%
Wouldn't Vote 3.0%

[Seriously?]

This is legitimately concerning.

Not really.

A 17 point swing in a week isn't concerning to you? 

It's a little more than a week, but this poll has been giving out wacky outlier results for a while now. It's only just begun to swing back to anywhere remotely close to reality. I am a bit puzzled by the RV/LV lean to the right on the 2-way ballot, but lean to the left on the 4-way ballot. It's almost like the sample isn't the same voters.