You do not remember the poll showing Clinton with a 12% lead over. Really? You either have a very short memory or a very selective one. You don't remember the back and forth that we had over that poll 2 years ago where you first tried to argue that this poll was not oversampled with Democrats and then a did a near 180 when you said all polls were oversampled. Does any of this jog your memory, Speed Racer or do I need to cite and pull up that thread?
You say Rasmussen has a poor history of predicting races, citing a 78%. Where did you get this figure? And why aren't you citing the prediction rate of the polls that you are relying on? No polling outfit is going to predict every race accurately. Regardless of what you think, Rasmussen Reports has a pretty good track record. Are they always right, no of course not, but show me who is.
The 538 that you like adjusts other polls, using what they call "house effects" to make their adjustments. They downgraded a Rasmussen poll showing a Trump approval rate of 50% down to 45% and only upgraded a Gallop poll showing Trump's approval rate of 43 to only 44%. Does this sound fair? I don't think so. Yet, this is the polling outfit that you prefer. https://fivethirtyeight.com/features...roval-ratings/
I found another current poll from the Hill, which is not exactly pro Trump, showing Trump with a 46% approval rate. https://thehill.com/business-a-lobby...6-percent-poll
You have consistently understated Trump's popularity in your posts. At least now I know where you get your information, from a polling outfit that adjusts other polls using what the call "house effects" to make their adjustments. And they (538 with Nate Silver) were wrong on their prediction of the last presidential election. But hearing you talk, its only Rasmussen who has a very poor history of predicting elections. Lol!
Originally Posted by NiceGuy53
I remember several discussions on polls 2 years ago but you seem to know better than I do exactly what I said. To me it is not important. Today is.
I supplied you with the link to FiveThirtyEight in my earlier post:
https://projects.fivethirtyeight.com...roval-ratings/
If you took the time to look at it you would see the dates of polls taken, the results of those polls, and a column titled "Grade", under which each pollster is given a grade based on their accuracy over the years. When you click on the grade, details are given.
I did not mention the "Adjusted" column at all. YouGov has Trump at 39% approval, Ipsos at 40%, Quinnipiac at 41%, Gallup at 42% without using the "Adjusted" column. The reason I prefer this site is that it gives a summary of ALL polls, and does not cherry pick one based on one's political leanings.
The survey by "The Hill" is another data point that should be taken into consideration when reviewing polls.
We've discussed the inaccuracy of the polls in 2016 several times, too many times in my opinion. Most pollsters got it wrong in 2016 for whatever reasons at the POTUS level. As I posted in a thread, Larry Sabato definitely missed on his projection for POTUS but was off by only 6 out of 435 in his predictions for the House in 2016.
I find it entertaining to discuss issues like this. My goal, unlike your's, is not to prove others wrong in their opinions but rather to discuss the issues.
