I believe that many statisticians would look at that graph and conclude that some trend could be established based on the data. I'm not interested in getting wound around the axle over this particular graph. If you want me to say that you might have found a factual error, then "congrats" you might have found a factual error.
What do you think MOE means?
Can you explain why margin of error doesn't actually mean margin of error in this case?
Or will you just avoid the question again and pretend I'm strawmanning or being pedantic or some other cop out that avoids saying anything of substance?
But that really seems tangential to what I *think* your bigger point is, and that seems to be about whether we should trust polls. And to that broader question my answer is: consider the source and the other characteristics you can evaluate. Some polls will be useful, other will not.
In this case, the fact that polls are much less reliable than claimed is totally irrelevant.
If the poll is
exactly as accurate as claimed your narrative is still false.
Why do you care about Trump telling not the truth, but are willing to spread misinformation yourself in the same thread?
Within the MOE, you cannot construct a narrative without being misleading. You can't be wilfully misleading at the same time as claiming a moral high ground regarding honesty and integrity.
we cannot let lies become normalized.
For me, we should not let people spreading BS narratives based on misrepresentation of data become normalised (well it is normalised, I just think we should stop it).
It's amazing how many 'rational' people disagree with this basic point.