Too close to call in the Kentucky Dem primary right now: Kentucky Senate Democratic primary between McGrath and Booker to decide who challenges McConnell too close to call
I posted too soon… four more senate polls released today.
U.S. Senate, Mich., JUN 17-20, 2020, Hodas & Associates*, Peters 51%, James 38%, Peters +14
U.S. Senate, N.C., JUN 14-17, 2020, Redfield & Wilton Strategies, Cunningham 45%, Tillis 36%, Cunningham +9
U.S. Senate, Ariz., JUN 14-17, 2020, Redfield & Wilton Strategies, Kelly 49%, McSally 34%, Kelly +15
U.S. Senate, Mich., JUN 14-16, 2020, Redfield & Wilton Strategies, Peters 50%, James 32%, Peters +18
Tillis running NC TV ads haven’t seen any for Cunningham but I may have missed them. the main way Tillis loses is if Trump loses here or just barely wins.
Bijou, are you saying the current polling doesn’t accurately reflect what those who will actually vote are thinking NOW — or, are you saying it’s likely that many will change their mind (and vote for Tillis after all) over the next several months — or, are you saying nearly all the current undecideds will end up voting for Tillis? Or some combination of these things?
This isn’t a “gotcha” — I’m genuinely curious.
There is no way Tillis only gets 36% in November so that poll is way off . Both guys will get around 45% to start. Debra Ross was a pretty weak candidate and she got 45% in 2016. Burr only got 51% to win. If Cunningham wins he will get 52% at most. Dem Kay Hagan got 52% in 2008 probably helped by Obama.
That poll is not trying to predict the share of the vote that Tillis will get in November. That’s not a thing polls do or how polls work.
Maybe it’s better for another thread, but if a poll like that is not trying to predict the likely vote share, then what IS it trying to do? If the answer is “give us a reasonable approximation of what the current electorate landscape is amongst a particular geographic area” (or something to that effect), then what’s the functional difference between those two?
Polls don’t predict elections? Where does that idea come from? Then why do people spend all the money on polls?
They surveyed 902 likely voters in North Carolina and asked them who they were going to vote for in November.
36% said Thom Tillis
45% said Cal Cunningham
3% said Other (Another Third Party/Write-In)
16% said Don’t Know
They then reported those results.
They are not predicting that 16% of the people who go to the polls in November will cast a vote for Don’t Know. That is ridiculous and nonsensical.
They are not predicting anything at all. They are just reporting the results of their survey.
Don’t all polls have a number of “don’t know” answers? Another way of looking at: as of right now only 36% of people say they are voting for Tillis, right?. Isn’t that kind of low for an incumbent ? Since 2000 the only Dem to win a NC Senate seat is Kay Hagan who only lasted 1 term. Only Obama in 2008 won NC for president since 1976. Those numbers are why I say Tillis is favored.
There’s some truth to that, sure. Basically, the third option I mentioned — trying to guess which way the “undecideds” will break in November. That certainly was a key to understanding Clinton-Trump ‘16.
If that were true, they would never bother with stating what the margin of error of a poll was. There’s no error in reporting how 902 people responded.
Do you think that this poll is predicting that 16% of voters plus or minus the margin of error are going to go to the polls on election day and vote for “Don’t Know”?
Or do you think that this poll is stating that 16% of the electorate plus or minus the margin of error don’t know who they are going to vote for in November at this point?
Then why bother with reporting this information at all? Why bother with a discussion about it?
Because we’re political junkies and, however flawed, this is the best info we can now get?
Right? I don’t even understand how these are questions.
Why do we collect data? To better understand the world.
Why do we discuss it? Because we find it interesting.
After going through the thread again, I see repeated use of these polls paired with language that could only be described as “predictive”. “It’s seeming less and less likely…” is a predictive interpretation of the polls. The polls are being used to predict. That’s obvious. Of course they are - that’s what they’re for. Why push back on that? I honestly don’t understand the need to condescendingly suggest they’re just sterile pieces of information that we just look at, as if they’re in a museum. The questions being used in the polls are predictive themselves. “Who do you plan on voting for in November” is asking people to predict their own behavior.
Polls are data that can be used to predict the outcomes of elections. No one is refuting that.
What I was refuting was that this particular poll was predicting that Thom Tillis would get 36% of the vote in November. That’s not a proper interpretation of that poll or the concept of polling in general.
He posted, “There is no way Tillis only gets 36% in November so that poll is way off.” No one. Not that poll or anyone else is predicting that Thom Tillis will only get 36% of in the coming election.
That’s fair - and an important distinction to make on these early polls. I know I’m probably guilty of the same thing. But like Elendir’s Heir said - it’s the best information we have on hand.
As an aside, am I the only one who wishes that these poll aggregations would specify the parties of the candidates? I mean, I try to keep up with politics, but given a list of 100 Senators, I probably couldn’t tell you the party of most of them. And trying to track all of the candidates? Not a chance. When I see that Smith is 10 points ahead of Jones in North Takoma, what should I take away from that?