Search results
Results from the WOW.Com Content Network
Over last week or two her lead has shrunk by about 0.5 points according to 538 polls. Both nationally and in swing states. Some swing states by closer to 1 point. :-(This is contrary to the expected D convention bounce. I am especially worried about PA being so close. P.S.
Trump received slightly less than polls said in Iowa, NH and SC ... but did better than polls said in Nevada. Nevada will be an important swing state in the general election, while Iowa and South Carolina are clearly Republican and unimportant. NH is lean Biden. So, Trump did a bit worse than polls said in 3 states, better in 1.
As time passed, though, Rasmussen’s inability to meet the standards set by 538 — and two dubious polls conducted for right-wing organizations — eventually led 538 to make the change this week. MSN
But this thread, conversation and discrepancy with RCP has certainly increased my 538 curiosity and my curiosity about the discrepancy part. So, I just went into 538 and averaged the list of polls they show under their National Poll average (with Harris +3.6).
Well, it happens to be 538 itself who is most active in that pollster ranking arena and I bet you can certainly imagine how surprised I was that many of the pollsters 538 uses are nowhere to be found in 538's own pollster ratings and ranking list (which includes a whopping 277 ranked pollsters and an additional list of MANY more which are ...
Here is 538's commentary in a nutshell: Back in 2016, Hilary Clinton failed to get at least 50% in polls. She consistently polled in the mid-to-upper 40s. In other words, there were a large number of undecided voters and those voters ended up breaking toward Trump. When you look at RCP's finally polling averages, you saw that very thing: MICHIGAN
So back to the average page on 538, these 3 pollsters I discussed account for 8 of the 18 polls which appear to create the 538 average. These 8 polls are on the high end of favorable to Harris and all 18 on the page have Harris in some sort of a lead over Trump. I'm curious as to what has led you to the conclusion "they are more reliable than RCP"?
538 dropped Rasmussen from their pool of polls because they found it unreliable. Rasmussen also has incredible volatility; successive polls will swing wildly. On a straight average (I assume RCP weights them, but I do not know), Rasmussen is the only one above showing strength for Trump.
This is an update to my earlier thread of 2020 vs 2024 polls. I had pointed out there that nationally in 2020 and in swing states, Biden was quite ahead of Trump according to 538 collection of polls. Yet, in the end, despite winning popular vote by many millions, Biden barely won in 2020 by just 44,000 votes in Georgia, Arizona and Wisconsin.
Priority goes likely voters, registered voters, and adults when multiple polls are available. The number of respondents polled would be the tiebreaker if two of the same type of poll are released during the same day. The 538 percentage model (if you click on the link) is then generated by using only the polls of that week.