393
Comments (21)
sorted by:
24
Wrexxis780 24 points ago +24 / -0

This is a pretty sick question for a poll. Holy hell.

17
deleted 17 points ago +17 / -0
11
deleted 11 points ago +11 / -0
10
TheStoneOfSisyphus 10 points ago +12 / -2

Isn't it pretty suspicious that they already had a poll ready for this ? How can you conduct something like this in a few hours ? How many could they have screened ?

8
deleted 8 points ago +8 / -0
6
SkyKirasagi 6 points ago +6 / -0

Unfortunately, your math is flawed, as there's no numbers for independents. The full formulas would be:

26=40x+14y+az 74=60x+86y+(100-a)z

Furthermore, if I try the same math with the depressed/not depressed number, I get 56.25% republican, 43.75% democrat.

1
deleted 1 point ago +1 / -0 (edited)
3
deleted 3 points ago +3 / -0
2
deleted 2 points ago +2 / -0
2
sdl5 2 points ago +2 / -0

Subtle clue as to potential Trump voters self-identified as Dems:

Around 13% to 25% are noted as agreeing they feel sad, upset, worried, depressed, scared.

In the politically and socially divisive hate-filled media propoganda age of the last 4 years, this is a fairly high percentage of Dem voters that are going to be markedly displeased with the tone and narrative of the media over this- a media they know full well veers Dem partisan.

A month before the Election...

2
deleted 2 points ago +2 / -0
2
Fizbin7 2 points ago +2 / -0

It is two equations with two unknowns.

Then he multiplies the first equation by 1.5 to get a term with the same 60x that appears in the other. Then subtracts that from the other equation to eliminate one variable, the x. That gives us one equation for y alone, easily solved. Substitute it back into either to find the other unknown, x.

2
EagleI 2 points ago +2 / -0

I'm confused about the "very well" vs "somewhat well" wording.

0
deleted 0 points ago +1 / -1
3
panthicc 3 points ago +3 / -0

I believe it's "how well" the following words describe how they're feeling. So does "sad" describe how you're feeling "very well", "somewhat well", etc.

1
deleted 1 point ago +1 / -0
1
Fizbin7 1 point ago +1 / -0

Math checks out...

1
SkyKirasagi 1 point ago +1 / -0

Actually, it doesn't. The data is incomplete, it doesn't give the responses of people who don't identify as either republican or democrat. As such, x+y doesn't add up to 100%. We need data on that third group, z, representing independents, and without it, we cannot solve for x and y, or for their ratio to each other.

I ran the formula as OP did, assuming that there were no independents, for every line on the chart. Surprised and confident show 73% democrat. Angry gives 67% republican. Without more information, it's impossible to draw any kind of conclusion as to party makeup.

1
Fizbin7 1 point ago +1 / -0

You’re being an obtuse pollster, his original inference was sound.

Anyone who is happy Trump has covid is an objective Demonrat at this time, even if their registration says R. Likewise for the reverse. There are no neutrals on election day.

1
SkyKirasagi 1 point ago +1 / -0

No, his inference was not sound. If his inference was sound, you could run the same equations on any of the responses and get at least similar results, some slight variation of a point or two because we’re only given nearest whole percent. But you get wildly varying results. His logic is fundamentally flawed, and either you didn’t understand the poll, or you don’t understand how to solve a system of linear equations.

1
zestanor 1 point ago +1 / -0

In a lot of the other cases it indicates the opposite. There's too much noise in this poll, like all polls, to infer something deeper.