B-b-but RCP says that they’re tied in LA Times and IBD. Y-you mean RCP is lying to us? Faint./s
Results based on survey of 791 likely voters conducted from 10/16 - 10/21. Margin of error: +/- 3.6 percentage points. Party identification breakdown: (Unweighted) 250 Democrats/263 Republicans/260 Independents; (Weighted) 284/224/264.
Don’t know what weighted and unweighted means but that 284/224/264 is 772 total. 284/772 = 37%. 224/772 is 29%. 264/772 is 34%. That looks like D+8, but I don’t know what weighted/unweighted means. The unweighted is R+something and no one thinks that is a correct turnout model.
Don’t even bother looking at the 2 way RCP average. Look at the 4 way and subtract around a 1 point off Clinton’s 4 way lead (LA Times doesn’t do a 4 way and so isn’t included). RCP 2 way is right rounded - both are tied rounded. In the 4 way, IBD has Trump up +2 which is the accurate way to look at it. It’s likely if the LA Times did poll for four way it would look similar. So 4 way on RCP is 5.3% Clinton - 1% to add in LA Times = 4.3% Clinton lead based on the average polls, which includes 3 clinton outliers of +9-12%.