Reuters/Ipsos Natl: Clinton +4

[Likely Voter]

1232 Registered Voters ( 7/16-7/20)


Clinton    40%    
Trump    36%    
Neither/Other    13%    
Wouldn't vote    3%    
Don't know/Refused    8%    


4 way
Clinton    39%    
Trump    35%    
Johnson    7%
Stein    3%
Neither/Other    5%    
Wouldn't vote    3%    
Don't know/Refused    8%    

 
http://big.assets.huffingtonpost.com/2016ReutersTrackingCorePolitical72016.pdf

[Likely Voter]

As I noted the week press releases from Reuters/Ipsos count as new polls with new samples. I am treating them like Huffpollster and RCP.  This poll is like NBC/Survey Monkey in that they announce new toplines weekly. 

There is a site where you can dive into daily data, but they don't release that data as new, also that data lags behind the official releases by a day.

So these polls will be done weekly.  If people want to dive into the daily data stuff I guess that should be done in the latest Reuters poll thread up until a new weekly PDF release comes from Reuters.

[Likely Voter]

For now let's leave it as the USC poll has its own daily tracking thread and the Reuters poll is posted weekly when they issue their press release. If someone wants to track the daily data from Reuters post in in the latest weekly poll but remember the daily data lags behind. In other words new data wont be out for two days.

One thing definitely don't want to see is people going to the Reuters site and taking snapshots and posting those as brand new polls in between the weekly official Reuters releases.

If we end up with a lot of daily tracking polls maybe we will reassess then with a single sticky.

[Likely Voter]

The weekly release sent out by Reuters and used by the polling sites is RV. If you want to see apples to apples comparison to that and the OP, then the RV filter (not the LV filter) should be applied. 

Today's RV numbers are...

(7/18-7/22)

2-way
Clinton    40%    (-)
Trump    36%    (-)

4 way
Clinton    38%    (-1)
Trump    36%    (+1)
Johnson    6%    (-1)
Stein    3%